The Method Research On Haze And Atmospheric Particulates Concentration Of Beijing-tianjin-hebei Region Based On MODIS Data

Haze is mainly caused by suspended particles in the atmosphere. The research on haze focused on the aspects such as weather causes, mechanism of pollution and human health. But there is little research on haze monitoring method. However, remote sensing technology commits to the inversion and validation of the Aerosol Optical Depth for long time. And the atmospheric particulate matter mass concentration is estimated by statistical relationship which is built by Aerosol Optical Depth and particle matter mass concentration. And the temporal and spatial distribution of particle concentration is quantitatively calculated. But the Aerosol Optical Depth inversion algorithm requires more the assumption conditions. The formula is very complex and is necessary lots of ground correction so that it can be applied to other studies. It is necessary to monitor the haze and estimate of atmospheric particulate matter concentration using remote sensing image characteristics.This paper is based on the remote sensing image features by MODIS LIB, MYD03 and ground observation data of 2014 October. There are two kinds of method to monitor and estimate the particulate matter mass concentration.(1)We adopt MODIS three bands RGB true color image synthesis to identification the distribution of haze, and established the multiple linear regression equation to estimate the atmospheric particulate matter concentration. (2)We used haze optimized transformation algorithm (HOT)method to estimate atmospheric particulate matter concentration which combined with daily the surface reflectance products MYD09GA. Finally, based on the above research, this paper made the simple analysis to the internal and external causes of haze pollution using digital elevation model (DEM), MYD08 AOD monthly mean data and meteorological data in Beijing-Tianjin-Hebei region.The results show that (1) MODIS RGB true color synthesis can identify and monitor haze distribution. The significant correlation confidence level (two tailed) between RGB reflectivity and the atmospheric particulate matter concentration was 0.001. The correlation coefficient of PM2.5 and PM10 were 0.68 and 0.65 respectively. The multiple linear regression equations of the fitting accuracy were 86.8% and 76.3%. (2) From HOT image extracted HOT values and the corresponding ground observation of atmospheric particulate matter mass concentration regression analysis showed that HOT and atmospheric particulate matter mass concentration had a significant correlation confidence level of 0.001. And the correlation coefficient of PM2.5 and PM10 was 0.435 and 0.44 respectively; the quadratic model accuracy was 76.8%and 87.1% respectively. (3)Comparison of two methods for analysis found that the accuracy of the two methods was almost and could achieve very good results, but the RGB method is belong to a statistical relationship and HOT is a method of research about the mechanism. And HOT is more suitable for popularization and Application. (4) Drawn from the analysis of the distribution of aerosol particles and the causes of haze: the AOD of Zhangjiakou and ChengDe city and its north region are less than 0.2.The AOD of northwest of Baoding and Shijiazhuang are between 0.2 and 0.3. And their south-east is between 0.4 and 0.6. AOD of Beijing north is between 0.4 to 0.5, in other parts of the Beijing-Tianjin-Hebei region AOD is between 0.5 and 0.8. Overall, AOD in southern Beijing-Tianjin-Hebei region is greater than the north, namely plain area higher than in the north side of the high altitude area.The AOD of Beijing-Tianjin-Hebei plains from north to south gradually reduced.There is wind circulation between mountains and plains of the Beijing-Tianjin-Hebei region valley. The northwest is relatively high above sea level.And the southeast wind conveying atmospheric particulate matter is blocked.These are easy to lead to the accumulation of atmospheric pollutants in Beijing-Tianjin-Hebei region.